I'm sure it's a long-observed
correlation, but corporate anti-union sentiment could easily explain much if not all of that effect. Try translating it to a situation like Germany, where AFAIU industrial labor (union and non-union) has thrived across the board, at least relative to the U.S. (U.S. has still has a strong industrial base in GDP terms, but even where unions are non existent its share of the labor market has been in a long decline much steeper than in places where unions are stronger generally.)
I'm looking for a legal source for reading the paper but not finding any. (e.g. not in JSTOR.) Can you say how they came to their conclusion--does it specifically show causality or is it simply an observational correlation that neatly matches predictions from a classical economic perspective? I guess in some ways I'm splitting hairs, because if U.S. corporate sentiment is inherently anti-union, then the logic behind investment strategies is simply their logic, whether it's a necessary in some universal sense or not. And if we're not going to peek behind the curtains, there's nothing else to say.
But it does matter when discussing what's possible for U.S. labor markets given sufficiently comprehensive cultural change. Racism is endemic and systemic, there are no simple solutions, yet we don't simply stand back and say, "it is what it is, there's nothing we can do about it". (Well, some people say that, but it's not generally considered an honest assessment.) Neither do we say that inducing change would lead to a more inefficient economy, even when there's some evidence showing policy-induced disruptions and dislocations with negative effects. The nearly universal assumption is that improved social equity (as a general matter; obviously we have trouble agreeing on the meaning of that nebulous concept, but few honestly reduce it to merely meaning "whatever the market produces") would result in a more efficient economy and more productivity, we just need to do a better job of it and work harder at threading the needle. It turns out that there are many "naturally" arising power disparities between employee and employer that result in rather inefficient labor system equilibriums, slavery and serfdom being the most obvious, if extreme, cases.